Home ARC Tunnel For NJ commuters, Gateway is no ARC

For NJ commuters, Gateway is no ARC

by Jeremy Steinemann

A proposed track map shows how the Gateway Tunnel would lead into Penn Station South. (Click the image to enlarge.)

Editor’s Note: With the announcement that Amtrak would seek funding for a proposed Gateway Tunnel from New Jersey to New York City, one-time Second Ave. Sagas guest writer Jeremy Steinemann started a new blog called Gateway Gab. He’s going to track the progress of the Gateway Tunnel there, and he is allowing me to post some of his analysis here as well.

Yesterday, New Jersey Governor Chris Christie viewed the Gateway announcement as a vindication of his decision to cancel ARC, but as Steinemann explains in the post below, Christie is off base. From the cost comparisons to the benefits to his state’s commuters, Christie and his criticisms shouldn’t enjoy a moment in the sun, and although Gateway would help alleviate the congestion into and out of New York City, it doesn’t have the same impact for commuters as ARC would have. What follows is Steinemann’s analysis.

As part of yesterday’s announcement of the proposed Amtrak Gateway Tunnel, Senator Frank Lautenberg’s (NJ-D) office has released a comparison of the three trans-hudson tunnel projects: Gateway, the now-defunct ARC tunnel, and the 7 train to Secaucus. This nifty chart is available at Lautenberg’s website, but take a look:

If anything, the chart is a testament to just how beneficial the ARC Tunnel promised to be, noting, for example, that ARC promised to connect commuters to the subway lines at Herald Square. The biggest difference between the two projects, however, are the compromises that Gateway requires of NJ Transit. For passengers on trains from Bergen, Passaic, Rockland and Orange Counties on the Main-Bergen and Pascack Valley lines, Gateway will not provide the long-promised, one-seat ride to Manhattan (at least not at first).

Furthermore, NJ Transit will have to coordinate operating control of the new tunnels. The current tunnels — technically known as the North River Tunnels — are controlled solely by Amtrak, whose trains take precedence over NJ Transit — wreaking havoc on the commuting schedule when an inter-city train is delayed. In contrast, ARC promised a set of tunnels operated solely by NJ Transit. The details of the Gateway operating arrangement are not clear, but improvements over the current arrangement are essential. Finally, as I noted yesterday, Gateway increases NJ Transit’s peak train capacity by only 13 additional trains, as opposed to ARC’s 24 additional trains.

The Gateway project, however, also includes a grab-bag of brand new items. Whereas ARC mentioned it as a possible future project, new access for Metro North at Penn Station is a full component of Gateway. In fact, a presentation on Lautenberg’s site promises capacity for six hourly trains on Metro North’s New Haven and Hudson Lines. The Harlem Line would not have access to the station.

In addition, Gateway’s new Penn Station South adds four new platforms serving seven tracks underground between 30th and 31st streets from 7th Ave. to just west of 8th Ave. Finally, Amtrak’s proposal also suggests extending the 7 train not westward but eastward from its future terminus at 34th St. and 10th Ave to Penn Station. It seems the cost of this extension is not factored into Gateway’s estimated price.

A critical benefit of Gateway, that is not being widely touted, is the system redundancy that the connectivity to Penn Station provides. The ARC tunnels were to be completely separate from the existing Penn tubes. If, for some reason, a train dies in one tunnel, as happens frequently now, the Gateway tunnels will provide a back-up. Due to the complicated nature of the train platforms underneath Moynihan station, however, it seems impossible for NJ Transit and Amtrak to utilize three of the four tunnels for peak-directional flow as the LIRR currently permits in its four tunnels under the East River.

The price comparisons between Gateway and ARC, reported at $13.5 billion and $10 billion respectively, are also misleading. The Gateway Tunnel also includes the Portal Bridge Replacement project, which ARC critics demanded should have been considered part of ARC’s total cost. Like ARC and the Portal Bridge projects, Gateway will double the right-of-way from Newark Penn Station to New York Penn Station from two to four tracks. Any changes to the connection between the Northeast Corridor and NJT’s Morris & Essex and Montclair Lines (located just west of the Portal Bridge) remain unclear at this time.

To view the full PDF presentation on Gateway, click here.

You may also like


Berk32 February 8, 2011 - 2:13 pm

Can’t say I think there is much of a point of extending the 7 to penn station.

I do wonder if they have considered making the Gateway tunnel double leveled like the 63rd St tunnel that carries both subway and the future LIRR… looks like the location works out on both side – could extend the 7 train that way to NJ.

BoerumHillScott February 8, 2011 - 2:23 pm

The 63rd street tunnel was prefabricated and lowered into a trench on the bottom of the east river, while the proposed Hudson tunnels will be constructed with tunnel boring machines due to the longer distance and deeper channel.

Al D February 8, 2011 - 5:04 pm

Extending the 7 to Penn would permit a ‘1 seat’ ride to the GCT area for NJT commuters. It would also, presuming the actual construction of a station at 41st & 10th, provide the same for workers on the Far West Side.

John-2 February 8, 2011 - 2:28 pm

The other important difference when you’re talking about funding in the $10-$13 billion range is the potential breadth of support — ARC’s terminal set-up meant that any benefit to people not in New York or New Jersey would be ancillary at best, by possibly speeding up Amtrak trains due to less need for the existing tubes to operate at maximum capacity. The new option that would allow the two new tubes at least some sort of East River connection via the southernmost LIRR tunnels means the new tubes can be beneficial as part of an overall high-speed rail corridor from Washington to Boston. The more federal funding that is sought, the more it helps to have House and Senate members in states besides New York and New Jersey who can see a benefit from a Hudson River tunnel project.

Bolwerk February 8, 2011 - 2:37 pm

Build the tracks to Penn, lose the Penn South, and have NJ Transit joint operate a few LIRR lines, such as Woodmere or Port Jefferson. Billions saved, capacity at Penn improved, region is interconnected better.

BBnet3000 February 8, 2011 - 4:45 pm

Seriously. When I look at the double track Embarcadero BART in San Francisco, compared to the 20 tracks of NYP, its amazing how much they get done during rush hour, and how NYP is such a glutton for platforms due to their old-timey railroad orientation.

That said, SF actually needs some sort of parallel line or secondary platform, when theres delays during rush hour, Embarcadero can reach dangerous levels of crowding on the platform very quickly. Penn doesnt have this problem because passengers typically arent waiting on the platform itself.

Al D February 8, 2011 - 5:07 pm

Forgive as I’m not a BART expert, but isn’t this an apples and oranges type comparison? I would think comparing BART to NYCT is more accurate?

mike February 8, 2011 - 7:57 pm

I don’t believe so. BART is the regional commuter rail network whereas Muni is the more local public transportation option in the city itself (Muni = NYCT). Although BART is still very different from LIRR/MNRR/NJ Transit, his points are valid and have been stated before on here.

BoerumHillScott February 8, 2011 - 8:02 pm

BART is almost always classified as a “Metro” or “Heavy Urban Rail,” the same as the NYC subway.

It is similar in design to the DC Metro or most other modern subway/heavy urban rail systems.

Muni is a streetcar/light rail system where different lines combine and go underground for Market Street.

Alon Levy February 9, 2011 - 1:29 am

BART has service patterns typical of an S-Bahn, track infrastructure typical of a subway, and station surroundings typical of American-style commuter rail. So on the one hand it’s primarily suburban, with multiple lines converging to create a single high-frequency urban line, just like the Munich S-Bahn. But unlike an S-Bahn, it’s entirely greenfield, and incompatible with mainline rail, raising the costs of everything. And the dominance of parking lot stations in the suburbs, with very little development, is inexcusable.

The Washington Metro is also somewhat of a subway/S-Bahn hybrid, with all the problems that entails, but at least it has gotten the development part mostly right.

Alon Levy February 9, 2011 - 1:32 am

And… just as I post, I see Al said the same thing at the same time. Only he explains it with analogies that are a lot more comprehensible to New York railfans.

al February 9, 2011 - 1:28 am

BART is more like a hybrid system. It has short stations distances and headways in the cities like a Metro. In the suburbs, the headways and station distances are longer. It shares some of the characteristics of the Paris RER.

There is also the fact that it is the “Bay Area Rapid Transit” that serves multiple independent cities and a multitude of suburbs.

You might be able to picture it if you think of a setup where all of the Bronx IRT lines ran deep into Westchester County, and ran only on the Lexington Ave Local in Manhattan. Or where the 60th St tunnel serviced ALL Queens Blvd and Astoria service, that also ran into Nassau County.

If the eastern end of the A/C to had been built out (and no additional tunnel into Manhattan) in the Second System, with lines running to the Nassau County Line, that would have been NYC version of the BART. It would share the infrequent service due to running multiple spurs onto one pair of track into lower Manhattan, and long station distances on the Rockaway Branch.

It is also somewhat similar to how Metro North RR and current LIRR run, in stretches during rush hr (I.E. Port Washington Branch). Finally, Park Ave used to have numerous local stations between GCT and 125th St along the outer tracks.

Anon256 February 10, 2011 - 3:47 am

While the Park Ave tunnel was built with local stations at 59th, 72nd, 86th and 110th, the stations at 59th and 72nd were never used in regular service. The stations at 86th and 110th saw a handful of trains a day (hourly or so) for a few decades, but this was more like the meager service provided today to local Metro-North stops in the Bronx than anything comparable to RER/BART/S-Bahn. Both stations were closed by 1910.

Eric F. February 8, 2011 - 2:39 pm

“Gateway increases NJ Transit’s peak train capacity by only 13 additional trains, as opposed to ARC’s 24 additional trains.”

That actually strikes me to be a very large number of trains! To make it simple, cut it back to 12 trains. That’s an additional train every 5 minutes for the 90-120 minutes of a.m. and p.m. rush. Someone will correct me, but I think NJ transit only runs 3 lines into NY. Those being NEC, the coast line and whatever midtown direct line runs in. Presumably, with the new trackage, you add in the Raritan Valley line, so now you have four. The ability to allocate that many trains per hour over 4 lines seems pretty solid to me.

“Gateway will double the right-of-way from Newark Penn Station to New York Penn Station from two to four tracks.”

That’s way too sweeping a statement. Take an Amtrak or NJT train out of NY. The single track segment is pretty brief. There are four tracks into Secaucus and at least that many into Newark. It’s hard to tell there who owns what, but there’s a ton of trackage between Harrison and Newark.

So actually, the problem you get into later –if this thing is ever built — is that you have many sections of the NEC that is only four tracks, and I wonder if you start to have severe capacity issues down the line later once you clear up the Manhattan approach.

lawhawk February 9, 2011 - 10:26 am

I’m a regular rider on the Bergen/Main line and my wife is one of the users of the Secaucus Transfer to get into midtown. I would dispute several aspects of the assessment about the ARC/Gateway project.

For starters, NJ Transit couldn’t afford to keep its existing rail schedule, so they cut back several train lines – reducing traffic into Hoboken and Penn Station. Off-peak trains are limited on the North Jersey train lines, and parking is the single biggest impediment to greater commuter train usage. Where is NJ Transit going to get the funds to operate 12 additional trains into Penn Station, let alone the 25 claimed under ARC (and Amtrak gets an additional 8-12 trains – so the total capacity is actually close to the same number as under ARC, but with the added benefits to HSR on the NEC).

Only within the past 18 months has parking been an option at Secaucus. Before then, Secaucus was a billion dollar boondoggle because the transfer traffic wasn’t there. It’s now utilized for additional traffic from the parking lot and those who would use the station to attend events at the Meadowlands via the spur line that opened last year.

Building a 1-seat ride would once again turn the Transfer into a boondoggle that costs NJ Transit more than it could bring in to support the operating costs and debt repayment that is an ongoing drain to the agency.

Then, there’s the issue of the ARC tunnel and its purported capacity. Where exactly were the trains to be stored when not in use during peak periods? For Penn Station, there was access to Sunnyside Yards, while there was no such access for ARC, which meant that the trains had to go somewhere (back into New Jersey?) that would reduce the actual capacity.

North Jersey riders aren’t getting a one-seat ride, but that means that Secaucus remains viable and not the boondoggle that it has been since the outset. NJ Transit, if it built the loop connector, would be cannibalizing ridership from Secaucus, and riders from North Jersey would still be faced with increased commute times (5-7 minutes each way) to Hoboken as has been the case since the track realignment and opening of Secaucus (most folks still ride the train to Hoboken for transfers to Lower Manhattan and Midtown below 34th Street).

Alon Levy February 10, 2011 - 2:59 am

What you’re saying could more briefly be summarized as “Secaucus may be a terrible transfer, but to bypass it would be to concede that it was in fact a boondoggle.”

Bear in mind, I personally think a one-seat ride is overrated, and Secaucus should be improved by reconfiguring it – i.e. no faregates, direct escalators from the Erie platforms to the NEC platforms, two island platforms and no side platforms on the NEC level, some attempt at timing the connections. People transfer at Jamaica all the time; there’s no reason Secaucus can’t be almost as good.

Andrew February 10, 2011 - 7:34 pm

Your proposals are mutually exclusive. There is no way an escalator can run from an island platform to a perpendicular island platform. An escalator from a side platform to an island platform is potentially feasible.

Timed connections are printed on the timetables. They’re not guaranteed – if your train is late arriving at Secaucus, you probably won’t catch the connection indicated in the timetable – but that’s to avoid propagating delays from one line to another.

Alon Levy February 11, 2011 - 12:15 am

Why can’t an escalator run between two island platforms?

Andrew February 11, 2011 - 6:28 am

Because they’re perpendicular. Whichever way you orient your escalators, they’re not going to clear one of the tracks, unless the island platforms are extremely wide (about as wide as the escalator is long).

That’s not a concern with side platforms – the escalator can run perpendicular to the track.

Bolwerk February 11, 2011 - 12:49 pm

Can’t you just go down lengthwise? Or is this really impossible at Secaucus? (I’ve never been there.)

Andrew February 11, 2011 - 5:02 pm

Lengthwise with respect to which level? For the third time, they’re perpendicular – lengthwise for one is widthwise for the other.

An L-shaped staircase might be able to fit – see the staircases between the NQR and L platforms at Union Square – but not an escalator.

Douglas John Bowen February 8, 2011 - 2:51 pm

Yes, indeed: For all its flaws, current and potential, “Gateway Tunnel” is not a 9-mile, stub-end “commuter” railroad paralleling, but separate from, the Northeast Corridor. This is a net plus and — somewhat ironically — brings us back to the issue of Access to the Region’s Core.

Alon Levy February 8, 2011 - 3:33 pm

Penn South is still a stub-end. Fortunately, it’s a stub-end that allows trains from either the new or the old tunnel to access both the new and the old tracks. Unfortunately, it’s still unnecessary infrastructure and still costs about five times what most Continental European cities would be spending on the same project.

Bolwerk February 8, 2011 - 3:00 pm

Moynihan Station really should be scrapped at this point. Another expensive monstrosity.

al February 9, 2011 - 12:08 am

Agreed. The Farley Post Office is a Landmark Building in its own right. Keep as much of the existing structure and build a station in and under it.

Nathanael February 10, 2011 - 5:36 am

Um, as an Amtrak intercity rider I’d like a decent station, rather than a hole in the ground. The Moynihan project was one way to provide that; there are others, such as building a proper headhouse over “Penn Station South”, or knocking down Madison Square Garden and building a station. Doing nothing is just stupid and pathetic.

Current NY Penn is the worst big-city station in the United States for intercity travelers, and I’m including such “gems” as Houston.

Bolwerk February 10, 2011 - 1:03 pm

As a frequent user of Amtrak myself, I would like a railroad that runs on time and a station that’s convenient to get into and out of. Let’s put aside the unavoidable problems with Penn Station’s location: just get people there, and get them into a subway or onto the street as quickly and easily as possible. The billion(s of?) $ being wasted on Mohnihan, which will in no way improve mobility,* could go to straightening tracks, improving signals, investing in constant tension catenary, electrifying commuter rail, or any number of worthy local transit projects.

* It arguably makes it worse. Moynihan is further from the east side of Manhattan than the current Penn setup.

Alon Levy February 8, 2011 - 3:05 pm

Come on, Ben, you’re not going to include “Has decent access to the East Side” and cost figures in your table?

Benjamin Kabak February 8, 2011 - 5:11 pm

Not my table! This one’s all Senator Lautenberg’s.

Alon Levy February 8, 2011 - 7:07 pm

Ah, sorry. I can’t say I’m surprised.

Gary Reilly February 8, 2011 - 3:15 pm

Now couple the news of this project proposal with the timing of Obama’s push for $52 billion in new HSR funding over 5 years.

I think this one is going to happen.

Bolwerk February 8, 2011 - 3:23 pm

Optimistic much? Republikans in general don’t like trains because they improve market efficiency, and teabaggers probably think trains bring devils in from inner cities to ravish their wimmin and steal their stereos!

Alon Levy February 8, 2011 - 3:35 pm

It could actually happen – Mica’s pro-HSR.

I don’t think there’s much point without first abolishing reforming FRA regulations and coming up with revised national standards that are in line with best world practice, but at least it could help shut the anti-CAHSR NIMBYs up.

Eric F. February 8, 2011 - 3:42 pm

“but at least it could help shut the anti-CAHSR NIMBYs up.”

Why would it do that? I don’t really see why expanding a currently heavily-used RoW and station would be persuasive to allow a new HSR line to run through one’s California town. Besides, NIMBYs’ arguments are ever-shifting anyway. NY can’t expand LIRR in the very train towns that depend on the line.

Alon Levy February 8, 2011 - 7:08 pm

A very large fraction of the money would go to California HSR, whose current problem is funding and not NIMBYism. For most of the way, the people along the line support construction because they want the jobs and the access to LA and the Bay Area. The NIMBYs are cocooned in the Bay Area demagoguing about how grade-separating the local commuter line and running high-speed trains on it at low speed will be the end of the world.

Al D February 8, 2011 - 5:14 pm

I wonder how Gov. Cheapsie is going to figure out how to kill this one.

BBnet3000 February 8, 2011 - 5:40 pm

BART functions somewhat like a subway in some places, but by and large its really commuter rail, just in a much more modern form than NJT/MNRR/LIRR.

Basically, my comparison between Embarcadero and NYP was meant to illustrate that by functioning more like a Metro/Subway, commuter rail can be a lot more space efficient. Its my understanding that many foreign systems (RER, S-Bahn) operate more in this fashion as well. Obviously the major difference between BART/S-Bahn and LIRR/MN will always be that whether they through-run or not, Penn and Grand Central will load and offload like Termini, because they are the only stations in the CBD. These other systems are commuter rail that put multiple stations within the CBD as well as through-running.

SEAN February 8, 2011 - 7:05 pm

In Center City Philadelphia you have 7 double ended commuter rail lines being funneled into a single row with 3 stations, Suburban, 30th Street & Gallery @ Market East Instead of the set up that we have here in NYC.

You can correct the issues we face with through running opperating agreements. Now with the new tunnels you can opperate trains from say Stamford to Montclair State. This has been proven to work as exampled by the the “train to the game” durring the last 2 football seasons.

BBnet3000 February 8, 2011 - 7:45 pm

this post was supposed to be a reply to Al D above, dunno what happened

jim February 8, 2011 - 6:00 pm

Actually for some NJT riders, this is ARC. The 20 tph currently is 8 NEC, 8 NJCL, 4 MEL. The 33 tph under Gateway will be 12 NEC, 12 NJCL, 7 or 8 MEL, 1 or 2 Raritan Valley. That’s the volume these lines would have carried under ARC. It’s the Main, Bergen and Pascack Valley lines whose riders have been shafted. Under ARC, they would have had the option of trains direct to midtown. Now they won’t. And I doubt if NJT will want to build the Secaucus Loop, for if it does, then it will have to decide which trains to take away from NEC, NJCL and MEL to create slots for Main, Bergen and Pascack Valley trains.

Woody February 8, 2011 - 6:23 pm

“. . . new access for Metro North at Penn Station is a full component of Gateway . . . capacity for six hourly trains on [the] New Haven and Hudson Lines.”

I count that as two Express HSR trains per hour shooting up to/down from Albany almost non-stop. Assuming that when the Empire Corridor gets upgraded and electrified, it would need two frequencies per hour each way.

(That overlooked and under-commented-on statistical study ‘High Speed Rail in America’ put out by the group America 2050 had a trainload of info nuggets, like the surprisingly high rank for this route as a candidate for HSR. It scored 19.29, the highest non-NEC city-pair in the region, ahead of NYC-Springfield at 19.0 and DC-Richmond at 18.31, for example.)

That could leave one train per hour on each of the New Haven and Hudson lines, if Lautenberg’s site means only six trains per hour, three each way (or two each if it is six trains per hour each way).

Not so many trains from MetroNorth after all, and probably not many commuters. Most riders could be intercity passengers heading to Penn Station to connect to Next Gen Acelas going to Boston or DC.

Frank B. February 8, 2011 - 9:19 pm

Who would’ve believed that Amtrak would actually save the day? And thank you New Jersey Senators!

Honestly, with track connections to the LIRR tracks, and Metro-North Access, this will hugely benefit New York as well. Now New York and New Jersey can benefit equally; hopefully Christie will use that thick head of his and actually work with others to get this done!

kvnbklyn February 8, 2011 - 9:25 pm

I think a few points are being missed by this post:

1. The 24 additional NJT trains per hour that ARC would have provided was only theoretical capacity. In reality, additional capital improvements to tracks, signals and other infrastructure on the NJ side would need to be made before that capacity could be utilized by actual trains – all for a lot more money. The 13 additional NJT trains per hour provided by Gateway would be possible without additional capital improvements outside the scope of the project (except possibly additional rolling stock if that’s not included in their current cost estimate).

2. One of the good things about demolishing an entire block to build new tracks rather than digging a cavern beneath the public street is the money that could come from the real estate development built over the tracks. I don’t think this is something to take lightly, especially since if the real estate deal is structured properly, it could provide an ongoing stream of revenue for Amtrak well into the future.

3. Any comparison made in order to judge the merits of the Gateway proposal shouldn’t be compared against the already deeply (pun intended) flawed ARC project, but against the much better Alt G proposal that would have connected the existing tracks at Penn Station to underutilized existing tracks at Grand Central. If a tunnel between Penn and GCT could be built for anything approaching the new extension to Penn south of 31st Street, then the tunnel should win out as it would provide NJT riders with a one-seat ride to East Midtown while also increasing through-put at Penn (since fewer passengers would be getting on and off there).

al February 10, 2011 - 12:53 am

They’re going to demolish an entire block and parts of 2 others. That is a nice piece of real estate, as Midtown South/Far West Side of Manhattan is one of the 2 main areas of development in the 2010’s. The other is Downtown Manhattan by WTC. That said, the Sunnyside Yards is a lot bigger. Whatever they do in these 2 places, they should go for development that generates large recurring revenue for capital projects through leases and rents.

Nathanael February 10, 2011 - 5:38 am

I absolutely approve of Amtrak control over the new tunnels, though it sure does suck for New Jersey; it’s the right thing to do in general.

Alt G is still the right idea, but this project actually is Alt G compatible, unlike NJT’s.

Is Gateway a victory for Gov. Christie? :: Second Ave. Sagas February 10, 2011 - 5:34 pm

[…] course, the truth is far from what Christie is promoting, and it’s worth it to revisit how the Gateway is not the ARC Tunnel. As Jeremy Steinemann noted here on Tuesday, the new tunnel is weighted toward interstate travel […]

Can the White House jumpstart the Gateway Tunnel? :: Second Ave. Sagas February 22, 2011 - 1:39 am

[…] « For NJ commuters, Gateway is no ARC Retail Thoughts: An Apple Store in Grand Central » Feb […]


Leave a Comment